DIYEFI.org Forum

I decided to make a footprint for the Cinch Modular Integrated Connector Enclosure, as it's rather attractively priced at Mouser and seems rugged. It's a bit specific on the PCB shape though, there's practically no room to rig off-shaped boards. There's also limitation zones for components and traces, so you're likely not able to install just any PCB into it.

Thus, to use the ModICE it will most likely need a hardware fork. There are few things to consider while at it too. The dual heatsink enclosure can support 12 FETs (or TO220's to be exact) and the case isn't ventilated, decision is needed what to keep and what not to. The casing has very high protection rating; from proofing against water jet to full and permanent submersion. Data lines should therefore be routed through the main connector so this benefit would not be compromised. On the plus side, the data terminals (CAN, USB, SD card bay)can be installed to a spot much easier to reach than they would be on the actual box. The biggest downside seems to be the lack of prototyping and modifying room inside the case, but if, and I hope when, CAN and whatever serial protocols are implemented, there will be lots of possibilities for external expansion.

It might be a good idea to wait a bit and see into which form the FreeEMS standard settles first though, and stick to more generally usable PCB's until that time though.

I used this datasheet in making the PCB base for the unit:

http://www.cinch.com/pdfs/1219441260-58 ... _rev_C.pdf

The module is like this:


It's simply the two connector banks displayed as A and B.

The footprint is like this:


The blue lines are for component height zones; the details are in the PDF. Component and trace exclusion zones are marked on top and bottom silk layers. There are slots for TO220 springs. There were no pin order preferences for the connectors, so I used a simple running numbering starting from top left and running down then right. No component should be closer than 2.54mm from any PCB edge. Maximum component height on the bottom layer is 2.54mm.

The libraries should not require any more work than extracting them to the work directory, open both EESchem and PCBNew and add the cinch.* -libraries. The schematics module has the footprint filter set so netlist association should be a bit easier too.

EDIT: old attachment deleted.

Nice work, just as an FYI, I have a couple python scripts I used for making things like the MCU and such connectors. Hmmm, I just check github both puma and DFH, I don't see those scrips. Hope I didn't loose them. I'll see if I can get them posted somewhere. They can be handy when making those symbols with many pins.

Do you know if there is a downloadable solid model for this connector or case? I can convert them to allow for a nice KICAD 3D picture. Sometimes a phone call can produce such items. I see a model was used to make the PDF's, so I'm sure there is one. I don't know how to get hold of it.

Cinch didn't offer 3D models right off the documentation page, peculiar; it's very commonplace these days.

This enclosure is one to really benefit from having one, given the space restrictions and TO220 clearance. I'm sure they will provide one when asked though. The problem is they might not be to keen to have it distributed.

I could do a model with SolidWorks though, and if you can convert the IGES to VRML it'd be excellent - I'm not too keen on working with Wings3d. It's a great 3D modeler but not really CAD oriented.

I just called them and left a message that I'm looking for a model of 581 01 60 005 the 60 pin version. We'll see if they come back with a model.

High Z might make little enough heat, that it's OK with a plastic case. Remember that plastic will conduct heat, just not as well as many metals. Also if heat is a problem, INJ can probably be remote as well, like IGN.

The heatsinked version has two aluminium rails for heat dissipation, and thinking stuff through again, the only really high heat element are the P&H drivers - having the drivers distributed on both sides it should disperse the heat quite adequately. There are even pressure clips to hold the FETs firmly against the heat sink.

I guess there's no limits afterall

Whoops, I had made a mistake in the TO220 compression spring slot locations plus it seems the springs are finicky about the FET placement. I updated the board with correct spring placement slots and zones for the TO's. Strangest thing though, the FET's (or whatever TO's) are placed over the component/trace exclusion zone.



Also, I had supplied a RAR archive instead of ZIP, now that's fixed too.

There are still 48- and 30-way large enclosures and 18/30/48-way small enclosures to do if anyone feels the need to fiddle with the component editor

Amazing. Just Amazing :-)

re temp: yes, hot circuits are linear regulators and P&H. The former should be solved in Puma, but is a thing to consider in other designs.

The PCB must 1.67mm tall, thats an important little thing.

Nice work Karry

Thanks for the compliments

I'm trying to figure out a way to use switching mode power (the thread's already posted), but it'll take a bit more sorting out to work properly.

I think the actual PCB may be no more than 1.50mm thick, the solder masks and silk screens add up surprisingly quick.

Just got the above mentioned cinch model I requested. It was apparently only for the connector, so I put in a request for the enclosure 581 01 60 031 with 2 heat sinks. I converted the connector to a .wrl for KICAD 3D so you can include the connector if you would like. You'd have to figure out the scale, and rotation and such. Kind of a pain in the arse, but doable.

Just finished a push, it's on github now under puma_jah.

Excellent! I'll check it out in the morning. Greatly appreciated!